MIDU: Enabling MIMO Full Duplex Ehsan Aryafar Amir Khojastepour - - PowerPoint PPT Presentation
MIDU: Enabling MIMO Full Duplex Ehsan Aryafar Amir Khojastepour - - PowerPoint PPT Presentation
MIDU: Enabling MIMO Full Duplex Ehsan Aryafar Amir Khojastepour Karthik Sundaresan Princeton NEC Labs NEC Labs Sampath Rangarajan Mung Chiang NEC Labs Princeton ACM MobiCom 2012 ACM MobiCom 2012 Background Background AP AP p 2 p 1 Current
Background Background
AP C t i l di h lf d l
p1 p2
AP
- Current wireless radios are half duplex
Amir Khojastepour NEC Laboratories America
Background Background
AP C t i l di h lf d l
p1 p2
AP
- Current wireless radios are half duplex
- Same band Full duplex is hard
Self interference is very high 75 dB for 15 dBm Tx power – Self interference is very high: ≈75 dB for 15 dBm Tx power – Transmitted signal is known self interference cancellation – Self interference can be significantly reduced by adding a g y y g cancellation circuit: e.g., a cancelling antenna
Amir Khojastepour NEC Laboratories America
Background Background
AP C t i l di h lf d l
p1 p2
AP
- Current wireless radios are half duplex
- Same band Full duplex is hard
Self interference is very high 75 dB for 15 dBm Tx power – Self interference is very high: ≈75 dB for 15 dBm Tx power – Self interference can be significantly cancelled by adding a cancellation circuit: e.g., a cancelling antenna
Can full duplex wireless double the capacity?
Amir Khojastepour NEC Laboratories America
Full Duplex vs. MIMO Full Duplex vs. MIMO
- Hardware complexity,
performance, size, cost metrics
Amir Khojastepour NEC Laboratories America
Full Duplex vs. MIMO Full Duplex vs. MIMO
- Hardware complexity,
performance, size, cost metrics
- Antenna Conserved (AC): Same
- Antenna Conserved (AC): Same
# antennas
Amir Khojastepour NEC Laboratories America
Full Duplex vs. MIMO Full Duplex vs. MIMO
- Hardware complexity,
performance, size, cost metrics
- Antenna Conserved (AC): Same
- Antenna Conserved (AC): Same
# antennas
- RF‐Chain Conserved (RC): Same
# chains
Amir Khojastepour NEC Laboratories America
Full Duplex vs. MIMO Full Duplex vs. MIMO
- Hardware complexity,
SI loss: 6 dB Ant Correlation: 0.1 performance, size, cost metrics
- Antenna Conserved (AC): Same
= 0.01
- Antenna Conserved (AC): Same
# antennas
- RF‐Chain Conserved (RC): Same
80 100 nel use) FD-RC HD FD-AC
# chains
- Significant FD gains in RC model
40 60 ty (bit/chann
Significant FD gains in RC model
- Limited FD gains with small #
antennas in AC model higher gains with more antennas
5 10 15 20 20 Capaci
gains with more antennas
Amir Khojastepour NEC Laboratories America
5 10 15 20 Number of Antennas
Full Duplex vs. MIMO Full Duplex vs. MIMO
- Hardware complexity,
SI loss: 6 dB Ant Correlation: 0.1
= 0.01
performance, cost metrics
- Antenna Conserved (AC): Same
80 100 nel use) FD-RC HD FD-AC
- Antenna Conserved (AC): Same
# antennas
- RF‐Chain Conserved (RC): Same
Regions of pronounced full duplex gains in b th RC d AC m d ls
40 60 ty (bit/chann
# chains
- Significant FD gains in RC model
both RC and AC models
5 10 15 20 20 Capaci
Significant FD gains in RC model
- Limited FD gains with small #
antennas in AC model high gains with more antennas
5 10 15 20 Number of Antennas
gains with more antennas
Amir Khojastepour NEC Laboratories America
Outline Outline
- Background
Background i f
- Design of MIDU
- Experimental Evaluation
- Conclusion
Amir Khojastepour NEC Laboratories America
MIDU: MImo full‐DUplex MIDU: MImo full DUplex
- Symmetric antenna
T1 R1 T'1
placement
1 1 1
d d d d π
Input Signal Input Signal
Amir Khojastepour NEC Laboratories America
MIDU: MImo full‐DUplex MIDU: MImo full DUplex
- Symmetric antenna
RX Chain
placement
- 2 level of antenna
R1
- 2‐level of antenna
cancellation
– Tx cancellation followed by Rx cancellation
T1 T'1
π
Rx cancellation – Proved in theory to have additive gains under imbalanced gains/phase or
R'1
imbalanced gains/phase or imprecise placement
π
Amir Khojastepour
TX Chain
MIDU: MImo full‐DUplex MIDU: MImo full DUplex
- Symmetric antenna
placement
- 2 level of antenna
R2 R3
- 2‐level of antenna
cancellation
– Tx cancellation followed by Rx cancellation
R1
Rx cancellation – Proved in theory to have additive gains under imbalanced gains/phase or
T1 T2 T3 T'1 T'2 T'3
imbalanced gains/phase or imprecise placement
l b l
R'1 R'2
- Easy scalability to MIMO
Amir Khojastepour NEC Laboratories America
R'3
Implementation Implementation
- WarpLab implementation
– Narrow‐band 625 KHz – Open space environment – MIDU + MU‐MIMO MIDU + MU MIMO
Virtex‐IV Pro FPGA
NEC Laboratories America Amir Khojastepour
Implementation Implementation
- WarpLab implementation
– Narrow‐band 625 KHz – Open space environment – MIDU + MU‐MIMO MIDU + MU MIMO
- Performance metric: SNR or the
di Sh i corresponding Shannon capacity
Virtex‐IV Pro FPGA
NEC Laboratories America Amir Khojastepour
Implementation Implementation
- WarpLab implementation
– Narrow‐band 625 KHz – Open space environment – MIDU + MU‐MIMO MIDU + MU MIMO
- Performance metric: SNR or the
di Sh i corresponding Shannon capacity
- Spectrum analyzer based
Virtex‐IV Pro FPGA
- Spectrum analyzer based
measurement or the SNR reported by WARP
NEC Laboratories America Amir Khojastepour
Experimental Evaluation Experimental Evaluation
- Feasibility
– Channel–distance relationship – Stability – Impact on far‐field users
R2 R3
Impact on far field users
- Cancellation
R1
– Single‐level – 2‐level and MIMO
T1 T2 T3 T'1 T'2 T'3
- Comparison with MIMO
– Single link Single cell
R'1 R'2
– Single cell
Amir Khojastepour NEC Laboratories America
R'3
Impact of MIDU on ld Far‐Field Users
- Issue: How does symmetric
y antenna placement impact the far‐field users?
Amir Khojastepour
Impact of MIDU on ld Far‐Field Users
- Issue: How does symmetric
14 13
y antenna placement impact the far‐field users?
3 4 15 12
TX
1 2 5 6 16 11 9 10 7 8 20 17
<3m>
20 19 18
Amir Khojastepour
Impact of MIDU on ld Far‐Field Users
- Issue: How does symmetric
y antenna placement impact the far‐field users?
- Achieved SNR can be up to 4
dB higher/lower
Amir Khojastepour
Inner Circle Outer Circle
Impact of MIDU on ld Far‐Field Users
- Issue: How does symmetric
y antenna placement impact the far‐field users?
- Achieved SNR can be up to 4
dB higher/lower I f fi ld t
- In far‐field antenna
cancellation has very limited effect due to signal scattering (fading) g ( g)
- Similar results hold for RX
cancellation
Amir Khojastepour
Inner Circle Outer Circle
Experimental Evaluation Experimental Evaluation
- Feasibility
– Channel–distance relationship – Stability – Impact on far‐field users
R2 R3
Impact on far field users
- Cancellation
R1
– Single‐level – 2‐level and MIMO
T1 T2 T3 T'1 T'2 T'3
- Comparison with MIMO
– Single link Single cell
R'1 R'2
– Single cell
Amir Khojastepour NEC Laboratories America
R'3
Cancellation Cancellation
- Issue: Is 2‐level cancellation
T1 R1 T'1
additive? Is MIDU scalable?
- Connect the receiver to a
1 1 1
d d
- Connect the receiver to a
spectrum analyzer d d π
Input Signal Input Signal
Amir Khojastepour NEC Laboratories America
Cancellation Cancellation
- Issue: Is 2‐level cancellation
additive? Is MIDU scalable?
- 22
30 dB cancellation on each
- 22 – 30 dB cancellation on each
level separately
- Cancellation remains relatively
unchanged with Tx power
Amir Khojastepour NEC Laboratories America
Cancellation Cancellation
- Issue: Is 2‐level cancellation
RX Chain
additive? Is MIDU scalable?
- Phase shifter on each path to
R1
Θ
- Phase shifter on each path to
handle insertion loss and delay
T1 T'1
Θ+π
R'1 TX Chain
Θ+π Θ
Amir Khojastepour NEC Laboratories America
Cancellation Cancellation
- Issue: Is 2‐level cancellation
additive? Is MIDU scalable?
- Phase shifter on each path to
- Phase shifter on each path to
handle insertion loss and delay
- RX cancellation on top of TX
cancellation is additive
Amir Khojastepour NEC Laboratories America
Cancellation Cancellation
- Issue: Is 2‐level cancellation
additive? Is MIDU scalable?
- Phase shifter on each path to
- Phase shifter on each path to
handle insertion loss and delay
- RX cancellation on top of TX
cancellation is additive
- 4 dB decrease in cancellation
for the first added pair, 5 dB with 3 total pairs with 3 total pairs
Amir Khojastepour NEC Laboratories America
Experimental Evaluation Experimental Evaluation
- Feasibility
– Channel–distance relationship – Stability – Impact on far‐field users
R2 R3
Impact on far field users
- Cancellation
R1
– Single‐level – 2‐level and MIMO
T1 T2 T3 T'1 T'2 T'3
- Comparison with MIMO
– Single link Single cell
R'1 R'2
– Single cell
Amir Khojastepour NEC Laboratories America
R'3
Comparison with MIMO Comparison with MIMO
- Compare MIDU to MU‐MIMO
R
– RF‐Chain conserved model – Multi‐user beamfoming/filtering for MU‐MIMO in each direction UL DL interference in MIDU
R1 R2 MIDU BS
– UL DL interference in MIDU
- Metric: Shannon capacity of the
measured SNR
R5 R6
measured SNR
R3 R4
2m Amir Khojastepour NEC Laboratories America
Comparison with MIMO Comparison with MIMO
- Compare MIDU to MU‐MIMO
M: #UL Streams N: #DL Streams – RF‐Chain conserved model – Multi‐user beamfoming/filtering for MU‐MIMO in each direction UL DL interference in MIDU – UL DL interference in MIDU
- Full duplex gains diminish as
the number of streams is scaled the number of streams is scaled
Amir Khojastepour NEC Laboratories America
Comparison with MIMO Comparison with MIMO
- Compare MIDU to MU‐MIMO
3 DL Streams Var UL Streams – RF‐Chain conserved model – Multi‐user beamfoming/filtering for MU‐MIMO in each direction UL DL interference in MIDU – UL DL interference in MIDU
- Full duplex gains diminish as
the number of streams is scaled the number of streams is scaled
- For maximum full duplex gains,
the number of streams the number of streams between UL and DL should be dis‐proportionate
Amir Khojastepour NEC Laboratories America
Comparison with MIMO Comparison with MIMO
- Compare MIDU to MU‐MIMO
3 DL Streams Var UL Streams – RF‐Chain conserved model – Multi‐user beamfoming/filtering for MU‐MIMO in each direction UL DL interference in MIDU – UL DL interference in MIDU
- Full duplex gains diminish as
the number of streams is scaled
Full duplex has great potential in practical single cell MU-MIMO schemes in which the
the number of streams is scaled
- For maximum full duplex gains,
the number of streams
s ng c M M MO sch m s n wh ch th number of streams is small
the number of streams between UL and DL should be dis‐proportionate
Amir Khojastepour NEC Laboratories America
Related Work Related Work
- Single‐antenna full duplex
Single antenna full duplex
– M. Knox, “Self‐jamming for full duplex” Tx Signal Tx Output Enhanced Circulator design for full duplex wireless Enhanced Circulator design for full duplex wireless Tx Signal Tx Output Rx Inp t Interference Cancellation Circuit Interference Cancellation Circuit Rx Input
+
Amir Khojastepour NEC Laboratories America
Cancellation Circuit Cancellation Circuit Rx Signal
Related Work Related Work
- Single‐antenna full duplex
Single antenna full duplex
– M. Knox, “Self‐jamming for full duplex”
- Asymmetric Antenna cancellation
- Asymmetric Antenna cancellation
– J. Choi, et. al., “Achieving single channel full duplex”
Amir Khojastepour NEC Laboratories America
Related Work Related Work
- Single‐antenna full duplex
Single antenna full duplex
– M. Knox, “Self‐jamming for full duplex”
- Asymmetric Antenna cancellation
- Asymmetric Antenna cancellation
– J. Choi, et. al., “Achieving single channel full duplex”
- Analogue cancellation
- Analogue cancellation
– M. Jain, et. al., “Practical full duplex” – M Durate et al “Full duplex with off‐the‐shelf radios" – M. Durate, et. al., Full duplex with off‐the‐shelf radios
Amir Khojastepour NEC Laboratories America
Related Work Related Work
- Single‐antenna full duplex
Single antenna full duplex
– M. Knox, “Self‐jamming for full duplex”
- Asymmetric Antenna cancellation
- Asymmetric Antenna cancellation
– J. Choi, et. al., “Achieving single channel full duplex”
- Analogue cancellation
- Analogue cancellation
– M. Jain, et. al., “Practical full duplex” – M Durate et al “Full duplex with off‐the‐shelf radios" – M. Durate, et. al., Full duplex with off‐the‐shelf radios Our work presents the design and implementation f th fi t MIMO f ll d l t
Amir Khojastepour NEC Laboratories America
- f the first MIMO full duplex system
In Summary In Summary
- Designed and implemented MIDU, the first MIMO full duplex
i l t wireless system
- Enabled two stages of antenna cancellation with additive gains that
provided as high as 45 dB self interference cancellation provided as high as 45 dB self‐interference cancellation
- Built a prototype of MIDU with joint operation of 3x3 MIMO + Full
Duplex in practice Duplex in practice
- Provided guidelines for the design of an efficient MAC for single
cells employing MIDU nodes cells employing MIDU nodes
NEC Labs: http://www.nec‐labs.com/ Princeton EdgeLab: http://scenic.princeton.edu/ Princeton EdgeLab: http://scenic.princeton.edu/
Amir Khojastepour NEC Laboratories America